issues
search
unslothai
/
unsloth
Finetune Llama 3.2, Mistral, Phi, Qwen 2.5 & Gemma LLMs 2-5x faster with 80% less memory
https://unsloth.ai
Apache License 2.0
18.41k
stars
1.29k
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Fix DPO, ORPO
#1177
danielhanchen
closed
4 weeks ago
2
Unsloth full finetune: Does the fast speed and small memory come with a cost of performance degrading or not?
#1176
fzyzcjy
opened
4 weeks ago
6
torch.compile fails
#1175
fzyzcjy
closed
4 weeks ago
3
Fix/kaggle pytorch
#1174
Erland366
closed
4 weeks ago
0
[FIXED] Kaggle broken
#1173
danielhanchen
closed
4 weeks ago
2
[FIXED] `TypeError: 'NoneType' object is not callable`
#1172
BlackWyvernX
closed
4 weeks ago
11
Fix/patch tokenizer
#1171
Erland366
closed
1 month ago
0
A bug in save.py
#1170
serendipity800
opened
1 month ago
2
Gradient accumulation fix produces a crash during fine-tuning.
#1169
gj414c
opened
1 month ago
3
Will unsloth's default padding cause problem?
#1168
fzyzcjy
closed
4 weeks ago
4
How to use this as the reference policy?
#1167
serendipity800
opened
1 month ago
1
chore: update chat_templates.py
#1166
eltociear
closed
1 month ago
0
Windows installation guide in README
#1165
timothelaborie
closed
1 month ago
1
OOM during saving to GGUF after training
#1164
Frank995
opened
1 month ago
3
Gradient accumulation fix does change the max_steps value
#1163
tristan279
opened
1 month ago
1
Many bug fixes
#1162
danielhanchen
closed
1 month ago
0
feat: add support for multiple column shareGPT
#1161
Erland366
opened
1 month ago
0
AMD unsloth/kernels/rms_layernorm.py":22:0): error: unsupported target: 'gfx906' > RuntimeError: PassManager::run failed
#1160
unclemusclez
opened
1 month ago
2
Excessive disk space consumption? How much is required?
#1159
LostRuins
opened
1 month ago
2
Support CPU offload?
#1158
fzyzcjy
opened
1 month ago
10
[Error] Some tensors share memory, this will lead to duplicate memory
#1157
katopz
opened
1 month ago
2
Orpo-Trainer is not compatible with new gradient-acum changes
#1156
Nazzaroth2
closed
3 weeks ago
7
Different batch size (1,2,4), same training speed
#1155
fzyzcjy
opened
1 month ago
4
"FlashAttention only support fp16 and bf16 data type" error when using dora
#1154
nguyentd01
opened
1 month ago
3
[FIXED] `wandb: WARNING The run_name`
#1153
jhangmez
closed
4 weeks ago
2
ModuleNotFoundError: No module named 'huggingface_hub.utils._token'
#1152
nguyentd01
closed
1 month ago
1
fix: compute_loss bug
#1151
vo1d-ai
closed
1 month ago
1
Pip installation failing due to unsloth-zoo?
#1150
BramVanroy
closed
1 month ago
5
any plans to support vision models?
#1149
reza8iucs
closed
19 hours ago
3
[FIXED] `huggingface_hub.utils._token`
#1148
tristan279
closed
4 weeks ago
7
Training works, Validation Fails OOM (With Reproduction Notebook)
#1147
tommedema
opened
1 month ago
4
Gradient Accumulation Fix
#1146
danielhanchen
closed
1 month ago
0
About UnslothTrainer
#1145
Vital1162
opened
1 month ago
6
Nightly transformers breaks Unsloth
#1144
CurtiusSimplus
opened
1 month ago
60
ModuleNotFoundError: No module named 'triton.common'
#1143
alansmithee-cpu
closed
1 month ago
2
train with unsloth single machine multi-GPU card
#1142
jiapengfei12356
opened
1 month ago
2
Cannot find -lcuda while Infering
#1141
RickoNoNo3
opened
1 month ago
2
unsloth_trainer shows the wrong number of steps with 10x longer completion time
#1140
davedgd
opened
1 month ago
5
.CANT LOAD LLAMA 3.1 70B due to ValueError: Some modules are dispatched on the CPU or the disk.
#1139
MuhammadBilal848
closed
1 month ago
4
create conda env follow Conda Installation fail
#1138
lastrei
closed
1 month ago
2
Can not use unsloth on vphere with ubuntu vm (vGPU)
#1137
NeilL0412
opened
1 month ago
4
Dependency Conflict When Installing Unsloth with Torch 2.2.0
#1136
signocob2
opened
1 month ago
7
Leftovers of a chat template. (Ollama, Llama 3.2)
#1135
IDiMooo
closed
1 month ago
4
Gradient Accumulation Fix
#1134
danielhanchen
closed
1 month ago
0
Remove extraneous f prefixes
#1133
esadek
closed
1 month ago
1
Fine tuning without GPU?
#1132
TechieHustle
opened
1 month ago
2
Load qwen2.5-32b on 4 gpu
#1131
luoruijie
closed
1 month ago
2
Could not do sft on qwen 2
#1130
Naozumi520
opened
1 month ago
5
Using Unsloth fine tuned model in transformers.js
#1129
djannot
opened
1 month ago
3
Unsloth: Most labels in your dataset are -100. Training losses will be 0.
#1128
Hasan-Demez
closed
1 month ago
2
Previous
Next